Use this page as your “best answer” compass: logic integrity → measurable status → governed change.
CPM basics (must-know)
- Critical path: the longest path through the network (drives the finish date).
- Total float: how much an activity can slip before it delays a milestone/end date.
- Near-critical path: small float + high risk; manage it like critical when uncertainty is high.
Dependencies & constraints (quick rules)
- Prefer logic links over hard constraints. Constraints hide problems and distort float.
- Relationship types: FS, SS, FF, SF (know what each implies).
- Leads/lags should represent real-world physics/hand-offs, not “schedule cosmetics.”
Resource constrained scheduling
- Resource calendars drive realism (availability beats optimism).
- Leveling/smoothing changes the critical path—re-check drivers after resource changes.
- Watch for “perfect plan, impossible staffing”: over-allocation is a schedule risk, not a rounding error.
Baselines & change control
- Baseline = your measurement reference. Treat baseline changes as governed events.
- Keep version history + change logs (who approved, what changed, impact).
- Don’t re-baseline to hide variance; re-baseline to reflect approved scope/strategy changes.
Statusing (make data usable)
- Use a consistent cut-off date and definitions for percent complete/remaining duration.
- Validate reported progress against completion criteria and observable outputs.
- Garbage in → garbage out: bad status data makes every analysis wrong.
Schedule QA checks (high yield)
- Missing predecessors/successors (open ends).
- Excessive constraints or “date driving” instead of logic driving.
- Negative float and why it exists.
- Out-of-sequence work and broken logic.
- Unjustified lags, unusually long durations, and unrealistic calendars.
- Subcontractor schedules not aligned to the integrated master schedule.
EVM schedule metrics (core)
- SV = EV − PV (schedule variance)
- SPI = EV / PV (schedule performance index)
Interpretation matters more than memorizing formulas: what does the metric imply you should do next?